Back to All Projects
Human-Centered Safety Tool2026

Lodestar: GNSS flight-readiness briefing for drone operators

A crisis-resilient dashboard that turns five live data sources into one go / no-go picture

My Role

Concept, interaction design & development

Status

Working prototype, updated every 10 minutes from public data

Tools & Tech

ReactTypeScriptSGP4GitHub ActionsStatic JSONGDELTNOAA SWPCOpen-Meteo

Is GNSS reliable here for the next 6 hours?

The Problem

GNSS jamming, spoofing and degraded satellite geometry are growing risks for drone flights, especially in crisis regions. The information that explains them is scattered across aircraft data, orbital tools, space-weather bulletins, weather services and the news. A pilot would have to read all of them and combine them under time pressure.

The Idea

Lodestar is a briefing, not a map of raw data. You pick a location and get one readable answer for the next 6 hours, built from five checks, each with its own evidence so the answer can be questioned rather than trusted blindly.

How It Works

  1. STEP 1

    Interference

    Position-accuracy reports (NACp) from nearby aircraft on adsb.lol, aggregated into 1° cells over a rolling 4 hours.

  2. STEP 2

    Satellite geometry

    Orbital elements from CelesTrak and an SGP4 propagator run in the browser to predict geometry.

  3. STEP 3

    Space weather

    Live NOAA SWPC data, fetched in the browser.

  4. STEP 4

    Flight weather and incidents

    Open-Meteo forecasts, plus GDELT news reports optionally labelled by an AI model.

Design Decisions

No server to run

A GitHub Action crawls the public sources every 10 minutes, writes static JSON and publishes the site, which keeps it cheap and resilient.

A decision aid, clearly labelled

Thresholds are documented heuristics, not validated limits. Degraded GNSS is never presented as proof of jamming or spoofing.

Evidence over verdicts

Each of the five checks is visible, so an operator can see why the briefing says what it says.

Limits stated up front

Interference is inferred from aircraft, so coverage depends on air traffic and ADS-B receivers nearby. The monitored area is currently Europe, the Middle East and the Caucasus.

Where It Stands

  • A working briefing built from five independent public data sources
  • Automated data refresh and deployment with no backend
  • Documented sources, thresholds and limitations